[pulseaudio-discuss] internal speaker silent in Asus eee netbook (patch)

[Guenter Milde]

On Sun Oct 15 09:50:43 Tanu Kaskinen wrote:
> On Sat, 2017-10-14 at 18:12 +0200, Guenter Milde wrote:

I reported the problem to the alsa devs:
https://bugzilla.kernel.org/show_bug.cgi?id=197309

The following discussion is about a workaround in pulseaudio in case this
cannot be solved upstream.

> > > analog-output-speaker used to have "switch = mute" for Headphone, but
> > > that was changed in this commit:
> > > https://cgit.freedesktop.org/pulseaudio/pulseaudio/commit/?id=22aac4e9fdb3786178f7815a0cb2150f588b1582


> > I agree that generally, it is preferable to switch off the headphone
> > control if headphones are not plugged in. However, if the alternative
> > is not any sound from the speakers, muting the headphones (instead of
> > switching off) is still an acceptable workaround.

> I don't know what you mean by "muting the headphones (instead of
> switching off)". The alsa mixer switch only has two states. "Muted" and
> "switched off" mean the same thing.

I mean the following change in the path configuration:

--- /usr/share/pulseaudio/alsa-mixer/paths/analog-output-speaker.conf	2017-06-18 18:03:31.000000000 +0200
+++ /analog-output-speaker-only-mute.conf	2017-10-18 10:12:48.000000000 +0200
@@ -1,4 +1,4 @@
-# This file is part of PulseAudio.
+# This file is based on PulseAudio's analog-output-speaker.conf.
 #
 # PulseAudio is free software; you can redistribute it and/or modify
 # it under the terms of the GNU Lesser General Public License as
@@ -13,7 +13,9 @@
 # You should have received a copy of the GNU Lesser General Public License
 # along with PulseAudio; if not, see <http://www.gnu.org/licenses/>.
 
-; Path for mixers that have a 'Speaker' control
+; Path for devices where the speaker does not work if the 'Headphone' control
+; is turned off.
+; Workaround for (some?) Asus eeepc desktops with Intel NM10/ICH7 Audio Controller
 ;
 ; See analog-output.conf.common for an explanation on the directives
 
@@ -85,8 +87,8 @@
 ; This profile path is intended to control the speaker, let's mute headphones
 ; else there will be a spike when plugging in headphones
 [Element Headphone]
-switch = off
-volume = off
+switch = mute
+volume = zero
 
 [Element Headphone2]
 switch = off

Together with an eeepc.conf alsa-mixer profile set that selects this
configuration instead of paths/analog-output-speaker.conf and udev rules
that select the eeepc.conf profile set I get an acceptable workaround for my
EeePCs.
 
> > My test shows that with "switch=mute", I can toggle the mute-state of
> > speaker and headphones independently. Unmuting the speaker does not
> > unmute the headphones and vice versa.

> How did you test? Since the Headphone element mutes both headphones and
> speakers, you can't mute headphones independently from speakers. If you
> mean that you can mute the sink in pulseaudio while using headphones,
> and the sink gets automatically unmuted when you switch to the speaker
> port, that's because PulseAudio stores the sink mute status separately
> for the two ports and the alsa mixer gets updated during a port switch.

With the above patch:  after plugging in headphones, sound is in the
headphones and the internal speaker is muted.

In the pulseaudio mixer, I open the tab Output. There, I can select the
active Port from a drop down list with the options Speaker and
Headphones. 

Changing the volume or muting in the GUI elements below affects the active
port only:

* If I switch the active port to speaker, I can unmute the speaker and
  control the volume. Now, both headphones and speakers emit audible sound.
  
* If I switch the active port to headphones, I can mute the headphones.
  This does *not* change the speaker setting in my case:
  Despite the headphones plugged in, the speaker emits sound at the previous
  volume and the headphone is mute.

Generally, while I can only change one output port at a time with the
pulseaudio mixer, these changes do not affect the other output port.

I can hear the spike in speaker-sound mentioned in the commit message to
https://cgit.freedesktop.org/pulseaudio/pulseaudio/commit/?id=22aac4e9fdb3786178f7815a0cb2150f588b1582
but this is IMO less severe than no sound at all from the internal speaker.